Re: aria-hidden (updates need to implementation guide or spec)

Birkir Gunnarsson <birkir.gunnarsson@deque.com> wrote on 03/23/2015 
04:58:37 AM:
> When aria-hidden="false" I believe this input should be announced,
> even if it has no accessible name.
> <div aria-hidden="false"
> <input type="hidden" value="3">
> </div>

I don't think focusable elements of any kind that are hidden, whether by 
CSS or hidden attribute,  should ever be visible to screen readers even if 
aria-hidden is false. 

Matt King
IBM Senior Technical Staff Member
I/T Chief Accessibility Strategist
IBM BT/CIO - Global Workforce and Web Process Enablement 
Phone: (503) 578-2329, Tie line: 731-7398
mattking@us.ibm.com

Birkir Gunnarsson <birkir.gunnarsson@deque.com> wrote on 03/23/2015 
04:58:37 AM:

> From: Birkir Gunnarsson <birkir.gunnarsson@deque.com>
> To: James Craig <jcraig@apple.com>, 
> Cc: Bryan Garaventa <bryan.garaventa@ssbbartgroup.com>, Joseph 
> Scheuhammer <clown@alum.mit.edu>, Joanmarie Diggs 
> <jdiggs@igalia.com>, Alexander Surkov <surkov.alexander@gmail.com>, 
> Dominic Mazzoni <dmazzoni@google.com>, "W3C WAI Protocols & Formats"
> <public-pfwg@w3.org>
> Date: 03/23/2015 04:59 AM
> Subject: Re: aria-hidden (updates need to implementation guide or spec)
> 
> Greetings.
> The confusion arises from two different source.
> First, there is the special case of <input type="hidden">
> When aria-hidden="false" I believe this input should be announced,
> even if it has no accessible name.
> <code>
> <div aria-hidden="false"
> <input type="hidden" value="3">
> </div>
> 
> Currently it does in at least one SR/OS combination and I am not sure
> if I should file it as a bug, or if I should tell developers not to
> use aria-hidden="false".
> Even if we use a CSS selector to tie display: block (or any other
> visible combination) to aria-hidden="false" we will come across this
> problem if input with type="hidden" is used.
> This issue could be solved by mapping <input type="hidden"> to
> aria-hidden="true" in the html to ARIA mapping, (currently it does not
> have an intro) and make sure that is consistently implemented as such
> by browsers.
> 
> The second confusion is when CSS is used to hide sections of child
> content of an element whose attribute is set to aria-hidden="false".
> I have seen this happen a few times, especially in tabpanels
> containing expandable regions or modal dialogs:
> 
> <code>
> <div role="tabpanel" id="foo2" aria-hidden="false" class="activeTab">
> ...
> <!-- expandable region inside the tab, initially in collapsed state... 
-->
> <button aria-expanded="false">Click to see invoice details</button>
> <div style="display: none;">
> Content that is hidden until user activates a button in tabpanel
> </div> <!-- hidden content -->
> </div> <!-- tabpanel -->
> 
> My understanding of the spec is that the invisible content inside the
> expandable region in the sample code should be visible in the DOM and
> announced by assistive technologies.
> I would personally prefer it if CSS determines visibility inside
> aria-hidden="false" container.
> This is not a problem if we remove the aria-hidden="true" attribute
> when content becomes visible, set its value to "undefined", or follow
> good coding practices and use aria-hidden to determine visibility for
> all users.
> Thanks
> -B
> 
> 
> On 3/20/15, James Craig <jcraig@apple.com> wrote:
> >
> >> On Mar 20, 2015, at 5:23 PM, Bryan Garaventa
> >> <bryan.garaventa@ssbbartgroup.com> wrote:
> >>
> >>>> The problem in this particular case is exacerbated by the
> >>>> recommendation that CSS selectors be used on the attribute to 
control
> >>>> visual display as it would encourage using aria-hidden="false" 
rather
> >>>> than removing the attribute when content is no longer hidden.
> >>>
> >>> Is this a hypothetical problem, or do you have a real-world example 
of
> >>> someone doing this?
> >>
> >> This is actually something that people have been recommending and
> >> implementing for quite a while now.
> >> " 5.  CSS selectors are used to bind aria-hidden and CSS 
display:none."
> >
> > Ah that. I thought Birkir was referring to something else.
> >
> > If display:none stays in sync with aria-hidden=false, as all your 
examples
> > do, there will never be a problem. We're just working on clarifying 
what
> > happens when they don't: both the authors expectations, and the UAs
> > implementation considerations.
> >
> > James
> >
> >
>